Evolving Tech, Emerging Complexity for Payment Integrity
Continuous glucose monitoring (CGM) is transforming diabetes care for the better - but with all transformations, these can create new billing or coverage challenges for plans.
Traditionally, glucose testing relied on finger sticks and manual tracking. Now, a new generation of CGM devices automatically measure glucose levels through tiny wearable sensors, transmit data to receivers or smartphones, and provide continuous, real-time insights for both patients and providers.
This innovation has expanded access and improved care, but it’s also introduced a maze of new billing rules due to new versions of care delivery but also an introduction of new medical equipment.
Modern CGM typically include three parts:
A sensor worn on the body, a transmitter that sends glucose readings, and a receiver (or compatible smart device) that displays data.
Each component can carry its own HCPCS code, yet CMS and the PDAC (Pricing, Data Analysis, and Coding) contractor have since established bundled monthly codes to capture the entire system (including all necessary sensors and supplies) under a single reimbursement rate.
For health plans, this coding evolution poses a growing integrity risk. When providers inadvertently continue billing daily or component-level codes, it can result in duplicated or higher overpayments. Recently, Shift uncovered this in action with Continuous Glucose Monitoring - finding more than $430,000 in potential overpayments for a plan.
Finding the Anomaly
In this case, Shift AI uncovered an unbundling trend where some providers were billing daily CGM codes in addition to separate sensor and receiver codes, rather than using the monthly bundled code defined by CMS and PDAC.
The pattern became clear. Providers were itemizing each CGM component such as sensors replaced weekly, transmitters every few months, receivers annually, and then layering additional daily monitoring codes on top. Individually, none of these services appeared improper, but the total effect was significant: across multiple billing cycles, unbundled charges led to an inflated reimbursement structure well above the CMS-defined amount.
By comparing provider behavior across peer groups, regions, and specialties, Shift’s algorithms flagged these outliers for expert review. What began as a subtle anomaly in code usage ultimately translated into roughly $850,000 in suspect claims, with $430,000 identified as recoverable once proper coding and fee schedules were applied.
Bringing External Data Into the Mix
Shift’s approach goes far beyond standard claims analytics. Our models integrate hundreds of external datasets, including PDAC listings, CMS fee schedules, and NPI registry data, to validate and qualify findings.
The PDAC dataset proved critical in this case. It maps each HCPCS code to specific CGM device models and effective dates, providing a definitive reference for compliant billing. By aligning provider submissions against PDAC guidance, Shift’s AI identified where daily codes were being used in place of the approved monthly bundles - precisely the type of error that editing systems can overlook.
Once identified, Shift’s certified coding experts validated the results. As Lisa Hornick, CPC, explains:
“We always strive to code to the greatest specificity while adhering to payer guidelines. PDAC is the gold standard for DME, and when plans deviate (even if that’s unintentional) the risk of unbundling and overpayment grows quickly.”
Human + AI: Expertise in the Loop
This case highlights Shift’s human-in-the-loop approach - where AI raises anomalies and experts validate them. Shift SMEs not only verify improper code combinations but also the differential between billed and appropriate reimbursement amounts. That enables health plans to act with confidence, supported by complete documentation and contextual evidence.
For payment integrity leaders, this process offers several key benefits:
- Targeted detection that uncovers complex, emerging patterns for upwards of 75% incremental findings
- Policy-aware validation that reduces false positives and provider abrasion in 60% of the time of manual processes
- Accurate financial outputs for defensible recoveries and savings attribution
From Detection to Prevention
Beyond recovery, the insight from this CGM case helps fuel and strengthen pre-payment detection logic. By integrating these learnings into Shift’s AI models, plans can proactively flag similar billing behaviors before payment - turning reactive identification into proactive prevention.
For payers, that means fewer leakage points, reduced rework, and a more transparent path to savings all grounded in explainable, defensible logic.
More Real-World Cases
This CGM analysis is one of several examples featured by Shift in our recent webinar, where our experts share real findings from live plan engagements.
Watch the webinar replay to see more or start a conversation to see how Shift’s AI-powered payment integrity platform helps uncover and prevent hidden overpayments before they happen.
